Assistant Professor
Dept. of Computer Science and Dept. of Statistical Sciences
University of Toronto
CIFAR AI Chair, Vector Institute
Research Scientist, DeepMind
cmaddis [at] cs [dot] [city] [dot] edu
Assistant Professor
Dept. of Computer Science and Dept. of Statistical Sciences
University of Toronto
CIFAR AI Chair, Vector Institute
Research Scientist, DeepMind
cmaddis [at] cs [dot] [city] [dot] edu
I develop machine learning algorithms with the long-term goal of advancing applications in the natural sciences. I am typically focused on improving our algorithmic tools, but I also enjoy collaborating on large-scale applied projects. I publish mostly at machine learning conferences (NeurIPS, ICML, ICLR).
The success of large language models is driven by the abundance and natural structure of big data. What does this tell us about our universe and ourselves? How can we use these insights to advance applications in other domains? I am interested in understanding how the statistical structure of real-world data influences the emergence of capabilities in these algorithms as they train on vast, heterogeneous datasets.
Large models and transfer learning Large language models like ChatGPT demonstrate that training large models on many inter-related tasks can have a synergistic effect. I am interested in understanding and applying these principles to improve machine learning in data-constrained settings.
Learning and optimization Algorithms for statistical inference and optimization are the engines that drive machine learning. Although inference and optimization may seem like distinct problems, there is a close interplay between them. I am interested in this interplay.
Applications in Scientific Discovery You can improve machine learning algorithms when you know something about the structure of the data. I have long been interested in applications that involve discrete reasoning. For example, I was a founding member of the AlphaGo project, which was the first computer program to defeat a world champion in the game of Go. I am now focused on AI for science and am launching collaborations to improve our scientific discovery tools.
Please check my Google Scholar for a complete list of my publications.
Prospective members If you would like to study for a graduate degree with me, you should apply through the CS department or the Statistics department. If you would like to work with me as a postdoctoral researcher, I encourage you to apply through the Vector Institute.
Current members
Here are some of my recorded talks, which cover the spectrum from academic talks to wistful reflections.